|
[Sponsors] |
September 30, 2013, 18:07 |
2d tri mesh prism layer
|
#1 |
Senior Member
Francisco
Join Date: Mar 2012
Location: Spain
Posts: 273
Rep Power: 15 |
Hi everybody,
I'm dealing with a pretty simple 2d tri mesh but I'm not able to get a prism layer around the central geometry. 1) Could anybody tell me what I'm doing wrong? 2) As it's a 2D case, I guess I have to treat my iner fluid domain as volume mesh, right? 3) Which algorithm should I use Octree or Delaunay? I attach case and geometry file. Thanks a lot! |
|
October 1, 2013, 09:24 |
|
#2 |
Senior Member
Francisco
Join Date: Mar 2012
Location: Spain
Posts: 273
Rep Power: 15 |
Searching a little more in the forum I've found two approaches to get a prism layer for a 2d tri mesh. I've tried both of them and none worked for me.
One approach is using the BLayer2D option in advanced prism meshing settings. The other is specifying the prism layers for the given curves. I describe here the steps I've followed for the 1st approach: 1 - Specify Global mesh parameters: global element scale factor and global element seed size. 2 - Set the Volume meshing parameters: mesh type (tetra/mixed) and mesh method (octree). I've been using volume mesh parameters even if my domain is 2D, in the end I compute the volume mesh instead of a surface mesh. Should I use surface parameters and compute the surface mesh? 3 - Set prism layer parameters and enable the BLayer2D option in advacned prism meshing. 4 - Set the surface mesh maximun sizes. 5 - In Compute Mesh/prism mesh select the curve and surface in which the prism layer is to be created. 6 - In Compute mesh/volume mesh check the prism mesh box. 7 - Compute volume mesh. As a result I get no prism layers in the selected curves. Can anybody tell me what I'm doing wrong? Is there a step missing? Can anybody tell me if I should work with surface rather than volume mesh for a 2d case? Thanks a lot you guys! |
|
October 1, 2013, 18:34 |
|
#3 |
Senior Member
Francisco
Join Date: Mar 2012
Location: Spain
Posts: 273
Rep Power: 15 |
I've tried to get a surface mesh instead of a volume one.
Setting the parameters to All tri, patch dependent gave horrible results. All tri patch independent have given a nice mesh (picture 1), but if I try to get a prism layer (activating the BLayer2D option) it gives strange results: 1) Part of the mesh disappear (blue elements in picture 1 are missing in picture 2) 2) Some shell elements are spread in a normal plane (picture 2) 3) There is a 3D prism layer in a created material part (picture 3). Also there is the corresponding 2D quad layers. How can I avoid missing all the shell elements and getting some of them in a normal plane? Any comment will be highly appreaciated. I attach case and tin file. |
|
Tags |
prism layers, tri mesh, unstructured mesh |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[ICEM] How can I create prism mesh for a 3D domain surrounded by a torus? | lzgwhy | ANSYS Meshing & Geometry | 5 | May 18, 2017 18:10 |
[snappyHexMesh] Layers don't fully surround surface | EVBUCF | OpenFOAM Meshing & Mesh Conversion | 14 | August 20, 2012 05:31 |
[ICEM] collapsed tri after prism mesh | skrzatchlewny | ANSYS Meshing & Geometry | 0 | January 20, 2012 05:08 |
Getting prism to inflate into mixed tet-hex meshes | Joe | CFX | 16 | October 10, 2011 08:06 |
[snappyHexMesh] snappy hangs while layer generation | bastil | OpenFOAM Meshing & Mesh Conversion | 3 | October 29, 2010 08:19 |